WeatherTileExamples

Simple examples to make tilesets for your favorite mapping system.

This is a simple set of examples that will show you how to take some of the newest, coolest NOAA products and turn them into tile sets for use in a web map mashup, or with WhirlyGlobe-Maply.

These examples rely on Vagrant and VirtualBox to run a Centos 7 environment which we will use yum to install gdal. Centos 7 with the epel gives us a reasonable install of gdal (version >= 1.11.2). We need this new of a version to handle opening the MRMS gz file without gzipping them (using a virtual file system) and for the alpha to work correctly with our palette.

About the data

Multi Radar Multi Sensor (MRMS)

MRMS is a software product built by the University of Oklahoma that is now being run in real time by the National Weather Service. The data can be retrieved by LDM or HTTP.

There are a number of interesting data products with MRMS including 3D slices of the atmosphere, lightning, hail, etc. We will show how to plot Reflectivity (Radar).

High Resolution Rapid Refresh (HRRR)

HRRR is a high resolution model that is run by NCEP and is updated very often. It can be found via NCEP NOMADS HTTP. We will look at the wind forecast, and composite reflectivity forecast (simulated Radar)

Steps to convert the data

1. Install Vagrant and VirtualBox
2. clone this repository
3. cd into the directory with the VagrantFile and type vagrant up
4. vagrant will load a centos image, and boot it. It will run the shell provisioner noaatiles.sh to install gdal, python bindings, and wgrib2
5. type vagrant ssh to get into the VM
6. type cd vagrant_data - this will put you in the scripts directory
7. type ./download_latest_hrrr.sh to download and create 3 tile sets from the latest HRRR model
8. type ./download_latest_mrms.sh to download and create a tile set from the latest MRMS

What do you end up with?

data directory

• MRMS and HRRR model files

output directory

• hrrr - a directory that has a tileset of the composite reflectivity plotted with alpha, that looks like future radar
• wind vectors split by component. Take a look at this page for more information: Wind: u and v Components
• hrrr_u_tiles - a tileset of the u vector of the wind
• hrrr_v_tiles - a tileset of the v vector of the wind
• hrrr_speed_tiles - a tileset of the speed of the wind
• mrms - a radar tileset from mrms data

Advanced modifications

More tile levels

Modify this line in create_hrrr_radar.sh:

   python ./gdal2tiles.py -r bilinear -z 0-9 output/hrrr.vrt output/hrrr

z is the number of z levels for the tiles. If you want it to create a higher resolution set of tiles, try this command instead.

####Change the Palettes This is inside palettes/radar_pal.txt

nv 		0 		255  	0 		0
0% 		0 		0 		0 		0
-30     255     255     255  0
-25      0       0       0   255
-20     255     255     255  255
-15     233     204     249  255
-10     207     128     223  255
-05     131     51      147  255
00      58       0      176  255
05      29       0      215  255
10       0       0      255  255
15      3       60      175  255
20      5       119     95  	255
25      8       179     15  	255
30      132     217     8  		255
35      255     255      0  	255
40      255     170      0  	255
45      255     85       0  	255
50      255      0       0  	255
55      179      0       0  	255
60      102      0       0  	255
65      51       0       0  	255
70       0       0       0  	255
75      250     197     250  	255
80      255     255     255  	255
100% 	255 	255 	255 	255